Avinash Kaushik Here, we talk to web analytics expert Avinash Kaushik about the future of web measurement, online surveys and ‘visitor experience optimisation’ – something he predicts will be a life or death thing for websites to adopt in the next couple of years.


A recent survey by the WAA suggested that many organisations are going to be integrating analytics with other applications this year in order to gain further insight into customer behaviour. What challenges do you expect them to face and what benefits are on offer?

One important note of caution - it is sexy cool to want to do global integration of all data in a corporation because at the end of that rainbow is a pot of gold. Or at least the promise of one. 

Please make sure first that you are really good at web analytics and are actively using data from your web analytics tool (be it Omniture or Google Analytics or IndexTools) to make improvements to your site.

There is a lot of low hanging fruit in those tools that is yummy and actionable. Consider this to be pre-requisite to future greatness.

Onto your question. No web team or website can live in a silo. You can optimise the channel, get very good at it but there will be value added insights you can gain from taking your web channel data and merging it with your phone channel data or your store data or your core ERP systems.

You can gain an understanding of cross channel customer behaviour; you can understand lifetime value better; you can identify high quality leads and so on.

There are two challenges to consider. Firstly, unless you are a bank website or an Amazon, both of which ‘force’ people to login, most of your web analytics data as it relates to ‘people’ is anonymous.

Best case scenario – your conversion rate is 2% and for that 2% you have name, address and so on. But not for most visits on your site. On the flip side, most of your other channels have well defined and identifiable PII data.

For your integration project consider what you need to be successful and if the data you need to join these different datasets exists (primary keys).

Secondly, you are going to need very flexible and new thinking when it comes to merging and storing all this data. In other offline channels traditional data warehouse mindsets and approaches work OK. But with your clickstream you will get a ton more data; a lot of it will be things you need to throw away and other parts of it will have holes. 

This needs a different mindset when it comes to designing your integration systems both from a design and data organisation perspective, and also from a hardware perspective (things like Netezza come into play for example).

As you evolve your business you will want to have a holistic view of your customers. Take promises of your consultants with a grain of salt, make sure you ask tough questions, cover the couple things above and jump!


Having just launched a free survey tool, what can companies gain by merging voice of customer insights with analytics findings?

Gleaning insights from your website data has meant a delightful and extensive torture of your clickstream data.

Sadly it sometimes comes up short because the story it tells is only about the ‘What’. Now, with free solutions like 4Q, you finally have easy access to the ‘Why’ data and hence the perfect marriage: What + Why = Less torture and more insights!

4Q to me is also about giving customers a real voice (not masquerading our opinions as our customers’).

Here’s an example. I am sure you looked at your “top ten pages viewed” report today to infer what people wanted from your website.

How do you know the top ten things that they actually wanted from your website? How do you know that? Worse, how do you know that when five of those ten things are not even on your site? The web analytics tool will come up short; you fill that hole with 4Q.


What developments do you expect to take place this year in terms of analytics APIs and partner programmes?

More and more of both. Data wants to be free and run around stark naked so you can stare at the insights even more clearly, so more API’s will be available. 🙂

But seriously though, the days of tool vendors being able to guess everything I could want or be able to build it for everyone are passe. There are so many people who are proficient at developing for corner cases and for the long tail.

More than anything companies will release APIs to take advantage of this fact.

In terms of partner programmes as well I think you’ll see more of ’em. I think partly this is fuelled by the fact that companies want someone ‘independent’ and not from their vendor to assist them and give them objective advice.

Partly it is because after struggling for so many years companies are waking up to the fact that the tool is not the be all and end all, massive success will spring from following the 10/90 rule for web analytics success: of the $100 you have, spend $10 on the tool and $90 on big brains (either in your company or get them from consulting companies).

Vendors are simply reacting to this need and creating such partner programmes.


Does your 90:10 rule still hold sway? Are there any situations in which this should be lower or higher?

Full disclosure first – while I created the rule a few years before I started consulting for Google, I want your readers here to note that I am a consultant for Google.

I think the rule applies even more today because if anything the web is evolving even more rapidly, organisations are becoming ever more complex, data capture on the web is even more of a challenge and finding the needle in the haystack is tough.

Tools are not the solution; people are. People are pretty and flexible and get marketing and can do analysis (not reporting!) and gather tribal knowledge and more mountains.

Invest in people who can make sense of the madness that is the web (and yes your organisation) and you’ll be rewarded with insights.

With tools like Google Analytics and Microsoft AdCenter Analytics getting better with each passing day (and the latter is good and still in a closed beta!), needing to spend lots on tools will be less of an issue.  Bonus reading material for your readers: Chris Anderson: Free! Why $0.00 Is the Future of Business.

There are some exceptions. As you move away from pure old school web analytics and on to Web Analytics 2.0, then maybe the rule is 15/85. There are some things still not free (though notice the survey just became free with 4Q).

Remember there is no exception to this today: Almost every expensive tool today (that will consume $90 of your $100 budget) will do a lot of complex things for you – that’s why it is expensive. It is inconceivable to imagine that it comes with auto everything.

You’ll need even more people to install and configure and run it. You’ll need people to analyse the complex data pouring out and find insights for you. You’ll need people to take advantage of all the advanced bells and whistles and so on.

There are at least 50,000 companies that need to buy these expensive tools and all 50,000 of them can benefit from the expensive tools. But not by simply buying it. They’ll have to invest a big chunk in people that will make the difference and get ROI.


Do you see more of a trend towards analytics for retention, not just acquisition and conversion?

Absolutely. It will take some time for us to get there but the stars are starting to line up.

One big challenge with retention always was how to know that you are a customer when you visit the site and what you have and how best to serve you in terms of the website experience. Yes we sent you emails and coupons and follow up notes. But that is primitive.

But now we have advanced behavioral targeting systems being made available for less than what it costs to buy the moon. These systems, even with anonymous data, are pretty good at knowing who you are and how to customise your experience (and it can be customised even more if you actually purchased something).

So technology is helping by making it easier to target. Additionally it is also helping because to create customised experience you don’t need to marry your IT person, you just get them to put tags or such and the targeting system takes care of the rest (freedom from IT clutches!).


What are your thoughts on site and content optimisation and the challenges and opportunities they offer organisations?

If you don’t have a roadmap that clearly outlines how you can do content optimisation (or as I call it: ‘visitor experience optimisation’) then you are going to be left with a dead website in a year or two. Because you’ll be the only one.

Be it through intelligent website platforms, such as ATG, or A/B; Multivariate Testing platforms such as the free Google Website Optimizer or paid tools such as Offermatica, or behavioural targeting platforms such as Kefta or TouchClarity, it is becoming easier to own and run optimisation technologies.

We are not quite at the ultimate nirvana of offering each visitor a custom experience, but you can certainly offer core segments of customers relevant experiences. How can acquisition / conversion / retention not improve?

The challenge a little bit is still price, but that will come down over time. The main challenge to me is having enough unique website content to be able to target it at all the visitor segments.

Most companies don’t understand enough about their customers (4Q to the rescue!) and don’t invest enough in creating content and creatives that will form the backbone of an experience optimisation system.

Remember if you input crap into the greatest optimisation system in the world it will only optimise crap, albeit it might do it really fast because it is advanced technology.


Can you describe your role at Google and how it’s going?

I am a consultant for Google, I am the Analytics Evangelist. I spend about 60% of my time at Google and the rest on strategic consulting with other companies and speaking engagements etc.

Much of my time at Google is spent with the various analytics teams at Google and helping make its customer-facing analytics products better.

Google is a unique place to work and I have had a lot of fun. I am struck by the raw opportunity to just get things done.

Hundreds of thousands of people use the tools and anything big or small that improves their ability to understand data better and make a smarter decision is a minor orgasmic experience. 🙂


Is Google 0% evil?

Yes. From time to time I run into people who disagree, but you have to be on the inside to be able to truly see what makes it tick (even if like me you are a contractor and not a full time employee!).

But even from the outside I encourage you to think for a moment what is the foundation on which Google’s modest success rests? Trust.

More specifically: customer trust. To be even 1% evil would be to jeopardise that most precious of commodities and I humbly believe that Google is a smart company.


Do you think that a more holistic view of online marketing may prompt companies to shift budgets away from search and onto other channels that help drive sales through search?

I fundamentally believe that every company should have a portfolio based acquisition strategy. So email, direct mail, catalogs, search, banners, social media etc etc all play a important role and it is every marketer’s job to ensure they are trying them all and investing in what works best for them in terms of ROI.

I am convinced that this will become even more deeply routed in how we all do business.


Related research:

Online Customer Engagement Report

Web Analytics Buyer’s Guide

Web Analytics Business Case Template File

Online Surveys and Research Buyer’s Guide

Related posts:

Interview: Jim Sterne, president of the Web Analytics Association